Deteksi Microaneurysm Pada Mata Sebagai Langkah Awal Untuk Penentuan Diabetic Retinophaty Menggunakan Pengolahan Citra Digital

Diabetic Retinopathy is a microvascular complication of diabetes mellitus. According to WHO (World Health Organization), there are more than 347 billion people who suffer from diabetes. This disease will become the seventh leading cause of death in the world in 2030. Based on research in Indonesia, it is estimated that there are 42.6% of diabetic retinopathy. Therefore, this final project plans a system to assist doctors in identifying diabetic retinopathy through its characteristics, namely microaneurysm. This system begins with an input retinal image from the fundus camera. Then the input will be processed in preprocessing to increase the contrast using the green channel. The next stage is segmentation. This is used to detect candidates from blood vessels and microaneurysms that use morphology operations. The next step is feature extraction, where it uses the features of glcm and white pixels detected in the image resulting from segmentation. The value of the white pixels and the values in the glcm feature are used as parameters in determining whether the classification process will be used as a determination of a Diabetic Retinopathy image or not. The success rate of the system using the SVM (Support Vector Machine) method is 88.4%.